the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
A diatom extension to the cGEnIE Earth system model – EcoGEnIE 1.1
Aaron A. Naidoo-Bagwell
Fanny M. Monteiro
Katharine R. Hendry
Scott Burgan
Jamie D. Wilson
Ben A. Ward
Andy Ridgwell
Daniel J. Conley
Abstract. We extend the ecological component (‘ECOGEM’) of the carbon-centric Grid Enabled Integrated Earth system model (‘cGEnIE’) to include a diatom functional group. ECOGEM represents plankton community dynamics via a spectrum of ecophysiological traits originally based on size and plankton food web (phyto- and zooplankton; EcoGEnIE 1.0), which we developed here to account for a diatom functional group (EcoGEnIE 1.1). We tuned EcoGEnIE 1.1, exploring a range of ecophysiological parameter values specific to phytoplankton, including diatom growth and survival (18 parameters over 250 runs) to fit best the model behaviour akin to observations of diatom biogeography, size class distribution, and global ocean biogeochemistry. This, in conjunction with a previously developed representation in the water column of opal dissolution and an updated iron cycle, produced an improved distribution of dissolved oxygen in the water column relative to the previous EcoGEnIE 1.0 as well as a value for global export production (7.5 Pg C yr−1) closer to previous estimates. Simulated diatom biogeography is characterised by larger size classes dominating at high latitudes, notably in the Southern Ocean, and smaller size classes dominating at lower latitudes. Overall, diatom biological productivity accounts for ~ 20 % of global carbon biomass in the model, with diatoms out-competing other phytoplankton functional groups when dissolved silica is available due to their faster maximum photosynthetic rates and reduced palatability to grazers. Adding a diatom functional group provides the cGEnIE Earth system model with an extended capability to explore ecological dynamics and their influence on ocean biogeochemistry through the late Mesozoic and Cenozoic, as well as enabling a broader range of palaeoceanographic proxies to be interpreted.
- Preprint
(3032 KB) - Metadata XML
-
Supplement
(659 KB) - BibTeX
- EndNote
Aaron A. Naidoo-Bagwell et al.
Status: final response (author comments only)
-
CEC1: 'Comment on egusphere-2022-1254', Juan Antonio Añel, 06 Apr 2023
Dear authors,
Unfortunately, after checking your manuscript, it has come to our attention that it does not comply with our "Code and Data Policy".
https://www.geoscientific-model-development.net/policies/code_and_data_policy.htmlTherefore, unless you make available the code that you use for your work in one of the repositories listed in our policy (the cGEnIE Earth system model and the different EcoGEnIE versions), we will have to reject your manuscript for publication in Geosc. Model Dev.
This preprint should have never been published in Discussions without the code available. In this way, I want to make clear that you must fix this in a prompt manner (without waiting until the end of the Discussions stage or replies to referees), as it is not possible to assess your work without the main part, which is the model you have developed. You must reply to this comment with the links to the repositories for the code and the corresponding DOIs.
Also, you must include in a potentially reviewed version of your manuscript the modified 'Code and Data Availability' section and the DOIs.Additionally, there are a couple of more issues that would be good to fix in your work:
- The PDF document that you cite in the Code Availability section of your manuscript reads at the beginning that the code is distributed "as is"and that this is the license that applies. However, the GitHub repository for Genie.muffin is released under the GPLv3 license. In this way, the statement that you include in the PDF manual is wrong, and you should modify it.
- EcoGEnIE 1.0 was published in 2018 in our journal (https://doi.org/10.5194/gmd-11-4241-2018). The "Code Availability" section for this paper is full of references to GitHub. However, GitHub is not a suitable repository for scientific publication. GitHub itself instructs authors to use other alternatives for long-term archival and publishing, such as Zenodo. Therefore, please, publish your code in one of the appropriate repositories according to our policy. Also, I would advise you to do not to include references to GitHub repositories, as they can generate confusion when looking for the version of the code used in your work. The Code and Data Availability sections are intended to comply with the principles of the scientific method about the replicability of your work, not to promote living repositories or projects.
Juan A. Añel
Geosci. Model Dev. Executive Editor
Citation: https://doi.org/10.5194/egusphere-2022-1254-CEC1 -
AC1: 'Reply on CEC1', Aaron Naidoo-Bagwell, 14 Apr 2023
Dear Editor,
Thank you for bringing this to our attention and apologies for any inconvenience caused.
The new DOI with all used code included is https://doi.org/10.5281/zenodo.7825320. The experimental configuration (including instructions for running the experiments) and the observational netCDF datasets used can be found in the following directory: genie-userconfig/EcoGEnIE1.1files.
The 'Code and Data availability' section will also be updated in the revised paper to reflect this change, and references to Github repositories and the PDF manual will be removed.
Thank you,
Aaron Naidoo-Bagwell and colleagues
Citation: https://doi.org/10.5194/egusphere-2022-1254-AC1
-
AC1: 'Reply on CEC1', Aaron Naidoo-Bagwell, 14 Apr 2023
-
RC1: 'Comment on egusphere-2022-1254', Anonymous Referee #1, 24 Apr 2023
The manuscript describes the addition of diatoms to the cGenie Earth system model. The diatom functional group is included as another functional group and the parameters were tuned by comparing with observed biogeochemical observations of dissolved oxygen, phosphate, and silicate. Overall, the methodology is well described, and the results clearly presented. The presentation of the diatom model could be improved to be more concise, as this model the main novel addition to the model (major point below). Several places the description of discrepancies were somewhat vague, and could be more precise (see several of the minor points below).
Major point: The paper revolves around the implementation of diatoms in the Genie Earth System Model. As such, the paper will be the future reference for the diatom sub-model. However, some aspects of the description are incomplete (see specific comments). A particular missing point is the grazing protection/palatability. Where does “phi_diatom” enter the model in equation (7) in the supplementary? And why is the “best run” value in table 2 (0.93) outside the tested range (0.3 – 0.8)? Finally, it would be nice to see a better presentation of the “trade-offs”, i.e., the differences between the diatoms and the pico/other phytoplankton, e.g., by graphs of the key parameters with cell size.
Minor points:
Section “Grazing” in supplementary. “The ’prey-switching’ term is optional …”. Please specify whether this term is applied in the present implementation?
Table S1: ”Diatom trade-offs”. It is a little unclear where the trade-offs refer to: higher Pmax than what? I suppose it is higher than the phytoplankton group. Another benefit is the higher maximum photosynthetic rate (should also be mentioned line 200).
Supplementary “temperature limitation”. What are the values of A and Tref? Are they the same as for the other phytoplankton group?
Line 162: ”and associated trade-offs”. Since the inclusion of diatoms is novel, it would be nice to referring to the trade-offs and dSi assimilation requirements already here.
Line 174: “We tested the implications of assuming the same 0.6µm to 1900µm across 8 size classes range in “EcoGEnIE 1.1y”. This sentence is unclear; what was the purpose of that test?
Line 188: “In contrast to EcoGEnIE 1.0, which applies a unimodal photosynthetic uptake rate relationship” How does this affect the model? It would have been nice if the implication of the different photosynthetic rates had been discussed in the interpretation of the results.
Line 194. It creates confusion to use the symbol “V” for both volume and max uptake rate.
Line 203: Please explain “bSi”
Lines 308-310: “The highest concentrations of phosphate (3 mmol P l-3) in the equatorial Indian ocean (2000 – 4000 m) seen in the data are limited to >2 km depths in the 310 model, likely due to restricted resolutions at depths and the smaller size of the Indian basin.”. It seems as if the vertical resolution is sufficiently fine to resolve more shallow high concentrations of P. Is it due to too low vertical mixing?
Eq (10). The equation number should be (3).
Table 2: Does the subscript “other” refer to the “eukaryotes” in table 1?
Fig.9 : It would have been nice to see Chl-a predictions with “EcoGEnIE 1.0” to assess how the inclusion of diatoms modifies this aspect of the model (if at all).
Figure 9b. I was surprised to see the relatively low Chl values in equatorial and eastern boundary upwelling systems. This does not fit well with the high concentrations of diatoms in these regions in figure 12d. It would be relevant to comment on this in the paragraph lines 329-.
Line 349. “opal”
Line 370: “Total model carbon biomass of all plankton resembles chlorophyll distribution with high values in the nutrient rich regions and low in the subtropical gyres (Fig 11b).” What about Southern Ocean? Chl-a is high at zones of the Southern Ocean (Fig.9b) where total carbon biomass is low. Why? Fig12c shows higher total diatom biomass in the Southern Ocean, but the total carbon biomass is lower here.
Line 417: “Despite a significantly lower global export in EcoGEnIE 1.1, specific regions have higher export relative to EcoGEnIE 1.0 (Fig. 15a).” Do you speculate somewhere the reason for this?
Lines 421-422: “That said, the absence of diatoms in NoDiatom intuitively results in a Southern Ocean with less export production than EcoGEnIE 1.1.” How does this compare to Fig11c at the areas of Southern Ocean where we see decrease in vertical POC fluxes in the diatom version?
Lines 437-438: “Equatorial chlorophyll biomass has a noticeable increase from the original rendition, producing results far closer to satellite estimates, likely a result of a thriving diatom population.” Do you show anywhere chlorophyll biomass from the original rendition? If yes, add reference here. Otherwise, it would be useful to include it.
Lines 487-488: “there are persistent limitations that ultimately arise from the efforts to trade off model complexity and user convenience”. This is quite a generic remark; some concrete arguments are missing here.
Fig.11b: Why does carbon biomass decrease in the southernmost cells of the Southern Ocean in the diatom version? Is it due to the sea ice module?
Figure 13 caption. “Eminence”? Do you mean “presence”?
Citation: https://doi.org/10.5194/egusphere-2022-1254-RC1 -
RC2: 'Comment on egusphere-2022-1254', Anonymous Referee #2, 19 Sep 2023
This manuscript presents a useful addition to the literature, and should in my view be published after addressing the relatively minor issues identified below.
Line 142: “The only differences with respect to the iron cycle parameterization used in Ward et al. (2018) are then: (1) the dust field of Albani et al. (2016) rather than Mahowald et al. (1999), (2) a mean global solubility of dust-delivered iron of 0.244 % as opposed to 0.201 % (partly due to the overall lower dust fluxes of Albani et al. (2016) vs Mahowald et al. (1999), and (3) a small reduction in the scavenging rate scaling (0.225 vs. 0.344 in Ward et al. (2018).” This is helpful information, but requires more justification. Presumably these numbers resulted from model turning, in which case this should be stated, and the objectives of that tuning described.
Line 151: “we do not attempt to calculate the fractional preservation of opal in accumulating sediments at the seafloor, but instead impose a simple benthic ‘closure’ term and reflect biogenic matter reaching the bottom of the ocean.” What do you mean by ‘reflect’ here? Please describe this more completely, what is the fate of this silica?
Line 235: What are you using old WOA data? Would updating this make a difference to your tuning, for example with increased data at high latitudes?
Line 245: “little further change occurred in biogeochemical indicators (oxygen, phosphate <1% change etc.)” Be more specific - max surface ocean concentrations, global ocean inventory….
Line 281: “Mean oxygen concentration produced by this iteration is also acceptable at 156 μmol kg-1, close to the 162 estimates” - please provide a citation for this estimate.
Line 286 I don’t think that “Overall, EcoGEnIE 1.1 captures the zonal contrast in phosphate concentrations between low and high latitudes” is a very good description. The contrast is in fact between the polar and sub polar regions, excluding the Arctic, and the the rest of the oceans. Low latitude implies latitudes around the equator, where the model performance is poor.
“The model-data comparison is also not strictly like-for-like, because in re-gridding higher vertical resolution WOA to the model grid, elevated subsurface concentrations become averaged into the re-gridded ‘surface’ layer.”. Does the physics of GENIE allow for a meaningful mixed layer to form? If not, there will not be the barrier to nutrients being entrained into the ~80m top level and the comparison seems reasonable? Further, can this argument apply to phosphate, but not to silica, where the results are pretty good, and one needs to instead justify why WOA is lower than your simulation in much of the ocean? Expanding your argument would help people like me who are not very familiar with the consequences of GENIE’s highly simplified physics.
“The spatial distribution of diatoms (all size classes combined) in EcoGEnIE 1.1 agrees with previous estimates (Tréguer et al., 2018)…” This is a key criteria for this manuscript, and needs to be expanded on. Currently I believe you only present results from your model here (please make figure captions more descriptive so this is not ambiguous) - this needs to be contrasted with other estimates or datasets in a more robust way, acknowledging the challenges around data availability. At present the verification of the distribution of the key new PFT you have added to the model is “… agrees with previous estimates (Tréguer et al., 2018), with high concentration in the productive regions (e.g. equatorial upwellings, subpolar regions) and peakings in the Southern Ocean at ~ 1 mmol C m” within the results section, and a comment on the relative size distribution in the discussion. Where observation based comparison is simply not possible because of limitations to available observations, explain this to the reader.
Line 377: “peakings” should be “peaking”
What hypotheses are stimulated by the trade offs seen in your M-scores? This behaviour is telling you something about the system or the limitation to the modelling approach - can you propose any suggestions?
General:
- ensure consistent formatting of references.
Minor specific points:
Line 51 “ration” to “ratio”
Line 598 sort out the reference and its citation in the text.
Table 1. Define ESD in the caption.
Citation: https://doi.org/10.5194/egusphere-2022-1254-RC2
Aaron A. Naidoo-Bagwell et al.
Model code and software
EcoGEnIE 1.1 configuration files Aaron Naidoo-Bagwell https://doi.org/10.5281/zenodo.7643548
Aaron A. Naidoo-Bagwell et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 421 | 131 | 18 | 570 | 57 | 6 | 6 |
- HTML: 421
- PDF: 131
- XML: 18
- Total: 570
- Supplement: 57
- BibTeX: 6
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1